The shapes of cities
Seen from above, a city tells its story: who laid it out, what it defended itself from, where the water ran. This guide goes through the main kinds of street pattern and where to see them best.
Reading a city from above
A street plan is a kind of fingerprint. Cities that grew on their own, house by house, have streets that follow the ground, old roads and field boundaries; those planned in one go carry the mark of a ruler and compass.
Almost no city is of a single type. The usual thing is an irregular old town surrounded by orderly extensions, cut through by later avenues and, further out, districts of blocks or detached houses. Those seams, where one pattern meets another, are often the most beautiful part of a map.
To see them you need the right scale. A shape 1 or 2 km across, such as a citadel or a walled town, looks best between 1:3,000 and 1:8,000; a grid neighbourhood or a whole city, between 1:15,000 and 1:30,000. The scale guide explains how to work it out for your paper.
The grid
It is the oldest of the planned patterns: the Greeks already used it (it is credited to Hippodamus of Miletus), and so did Roman camps, with their two main streets crossing, the cardo and the decumanus. Many European cities still keep that crossing beneath the medieval town.
The grid came back strongly at two moments: in the cities Spain founded in the Americas, around a main square, and in the nineteenth-century extensions. Barcelona's Eixample, by Ildefons Cerdà, is the most famous example: 113-metre blocks with their corners cut off so carriages could turn and light could get in.
In the United States the grid is almost the rule. Manhattan's was approved in 1811 and spread numbered streets across the whole island, hills included; San Francisco's climbs and drops impossible slopes without swerving. From above the difference with European ones is clear: elongated blocks and identical streets, with no square to organise them.
In a line drawing, a grid is an almost textile pattern. It works best when something breaks it: a diagonal, a park, a river or the coast.
Diagonals, axes and star squares
Diagonal avenues can be added to a grid, joining important points by the shortest route. Washington is the textbook case: in 1791 Pierre L'Enfant laid great diagonal avenues, named after states, over a grid of numbered and lettered streets, and made them converge on the Capitol and the White House.
Paris arrived at something similar by demolishing: in the mid nineteenth century Baron Haussmann cut straight boulevards through the medieval city and made them meet at squares such as the Étoile, where twelve avenues radiate from the Arc de Triomphe. On the map they look like the spokes of a wheel.
La Plata, in Argentina, takes the idea to the extreme: a square city, founded in 1882, with a perfect grid, two great diagonals crossing it corner to corner and smaller ones, and a square every six blocks.
Stars and polygons: fortress towns
When cannons arrived, tall thin walls stopped working. Italian Renaissance engineers invented the bastion: a pointed projection from which you could fire along the neighbouring wall, with no blind spots. A wall made of bastions, seen from above, is a star.
Some towns were built entirely on that logic. Palmanova, in Italy, founded by Venice in 1593, is a nine-pointed polygon with a hexagonal square at its centre from which streets run out like spokes. Naarden, in the Netherlands, has a double star of bastions surrounded by water-filled moats.
Others keep only the citadel: Copenhagen's Kastellet, a five-pointed star by the harbour, or the ramparts of Christianshavn across the water. They are some of the most rewarding maps there are: the shape is recognised instantly and fits whole on a small sheet.
Rings and circles
Many European cities grew in layers, like the rings of a tree. When the wall became too small, another was built further out; when walls were torn down, an avenue circling the centre was left in their place.
Vienna's Ringstrasse is the clearest example: in 1857 the walls were demolished and a horseshoe boulevard was laid out around the old town, closed by the Danube Canal, with the opera, parliament and city hall along it. Amsterdam did something similar with water: its seventeenth-century canals ring the centre in concentric arcs.
Rings call for a circular frame centred on the oldest part: the map becomes a target.
Mazes: the cities that grew on their own
At the other extreme are the old towns nobody laid out. The medinas of North Africa, the Moorish quarters of Andalusia or Europe's medieval centres have narrow, crooked, dead-end streets, because they grew house by house, following slopes, paths and the needs of each moment.
The Albaicín in Granada, the Jewish quarter of Córdoba or the centre of Toledo are good examples in Spain. Venice is a case of its own: a maze of alleys and canals over more than a hundred islands, in which almost nothing is straight except the Grand Canal, which crosses it in an S.
On the map, a maze is a fine, dense texture that calls for a close scale (1:5,000 to 1:12,000) and generous paper. Combined with something orderly around it, such as the wall or an avenue, it gains a lot.
Water and terrain: the shape the site imposes
Sometimes nobody decides the shape: the site does. Toledo is wrapped in a meander of the Tagus that surrounds it on three sides; Cádiz sits at the tip of a spit of land; Lisbon tumbles down hills to the estuary; Istanbul is split between two continents, on either side of the Bosphorus.
Rivers are the backbone of many maps. The Thames in London, the Danube in Budapest or the Vltava in Prague cut the city in two and give it direction. And some cities make their own shape out of water: Dubai's Palm Jumeirah is an artificial island shaped like a palm tree, with a trunk, seventeen fronds and a crescent that shields it from the waves.
For these cities it pays to frame the water, not avoid it: in a line drawing the sea and rivers are filled with shore lines and make a perfect counterpoint to the street pattern.
More guides
